AI should make everyday work easier. But a new study by the Boston Consulting Group of 1,500 employees shows a surprising downside. Those who intensively monitor AI systems suffer from significantly higher levels of mental fatigue, a phenomenon the researchers call “brain fry”. The article shows at what point AI use becomes counterproductive and why top performers are particularly at risk.
In recent years, artificial intelligence has rapidly developed from a niche topic to a key technology in many people’s everyday lives. AI is changing how people work, make decisions and even communicate.
In many companies, algorithms now automate routine tasks. This supposed gain in efficiency through the use of AI has also noticeably increased expectations of productivity and availability.
A new study shows that this can in turn have an impact on performance. Accordingly, certain usage patterns of AI lead to cognitive fatigue – a phenomenon that the researchers call “brain fry”.
Through the use of AI: How does “brain fry” occur – and who is hit hardest?
For their study, the researchers at the Boston Consulting Group surveyed around 1,500 full-time employees in the USA from large companies in various industries, functions and hierarchy levels about their everyday work with AI. This included, among other things, the patterns and extent of AI use, their work experiences and their cognitive and emotional stress.
The researchers were able to determine that the phenomenon of cognitive fatigue caused by intensive monitoring of AI systems is actually real. They call this phenomenon “AI Brain Fry” – i.e. AI-related brain exhaustion.
In this context, respondents described a feeling of buzzing or mental confusion. These in turn are accompanied by difficulty concentrating, slower decision-making and headaches.
For companies, such a development can bring exactly the opposite of the expected productivity gain. The AI-related mental stress can lead to increased errors and even decision fatigue and intention to quit.
Despite the observed effects, the researchers point out that mental fatigue is increasing. However, burnout levels decrease when AI takes over routine tasks. “This highlights the subtle but important difference between the types of stress that AI can alleviate and those that it can exacerbate,” it says Harvard Business Review. AI-supported workflows would therefore have to be designed in such a way that the risk of burnout is reduced.
14 percent more effort: Why AI monitoring is so exhausting
The survey results show a clear trend: For employees, direct monitoring of AI tools is the most cognitively stressful. The study participants who had a particularly high level of AI monitoring at work showed 14 percent more mental effort at work.
This was also accompanied by a twelve percent higher level of mental exhaustion. Additionally, information overload was 19 percent higher with more intensive AI monitoring.
Employee productivity is also influenced by the use of AI tools. The researchers observed that productivity increases when two AI tools are used instead of one. If a third tool is added, productivity increases again, but more slowly. With additional tools, the productivity values decrease again.
At the same time, the study shows a critical side effect of intensive AI use in the form of increasing intentions to quit. Those employees who use AI particularly frequently and are therefore often among the top performers in a company are at greater risk.
While around 25 percent of employees without significant mental exhaustion due to AI are actively thinking about changing jobs, this proportion is 34 percent among users who are heavily stressed. This corresponds to an increase in the intention to quit by 39 percent.
Also interesting: